Mehdi R. Zargham

A parallel Fuzzy C-Mean algorithm for image segmentation

This paper proposes a parallel Fuzzy C-Mean (FCM) algorithm for image segmentation. The sequential FCM algorithm is computationally intensive and has significant memory requirements. For many applications such as medical image segmentation and geographical image analysis that deal with large size images, sequential FCM is very slow. In our parallel FCM algorithm, dividing the computations among the processors and minimizing the need for accessing secondary storage, enhance the performance and efficiency of image segmentation task as compared to the sequential algorithm.

Parallel channel routing

A parallel algorithm is proposed for the problem of channel and switchbox routing in the design of VLSI chips. The algorithm is suitable for implementation on a shared-memory multiprocessor environment. The approach does not impose restrictions on the channel type (such as fixed or variable channel widths) and the number of available layers. The algorithm contains three major phases: (1) dividing the channel into several regions by selecting some columns, (2) assigning tracks to nets of the selected columns, and (3) assigning tracks to nets of the columns in each region.<<ETX>>

A self-organizing map and its modeling for discovering malignant network traffic

Model-based intrusion detection and knowledge discovery are combined to cluster and classify P2P botnet traffic and other malignant network activity by using a Self-Organizing Map (SOM) self-trained on denied Internet firewall log entries. The SOM analyzed new firewall log entries in a case study to classify similar network activity, and discovered previously unknown local P2P bot traffic and other security issues.

A Decision Tree-Based Classification Approach to Rule Extraction for Security Analysis

Stock selection rules are extensively utilized as the guideline to construct high performance stock portfolios. However, the predictive performance of the rules developed by some economic experts in the past has decreased dramatically for the current stock market. In this paper, C4.5 decision tree classification method was adopted to construct a model for stock prediction based on the fundamental stock data, from which a set of stock selection rules was derived. The experimental results showed that the generated rules have exceptional predictive performance. Moreover, it also demonstrated that the C4.5 decision tree classification model can work efficiently on the high noise stock data domain.

A Web-based information system for stock selection and evaluation

We have developed an expert system, called PORSEL (PORtfolio SELection system), which uses an effective set of rules to select and evaluate stocks on the Internet. At present, the PORSEL consists of three components: the information center, the fuzzy stock selector, and the portfolio constructor. The purpose of the information center is to provide representation of several technical indicators such as candlestick charts, moving average of closing prices, and price trends. The fuzzy stock selector evaluates the listed stocks and then assigns a composite score for each stock. The portfolio constructor generates the optimal portfolios for the selected stocks. A client/server model is implemented which allows users to communicate with the PORSEL program on a server computer at a remote site in a user-friendly manner. The results of simulation show that PORSEL outperformed the market almost every year during the testing period.

Vulnerability Scrying Method for Software Vulnerability Discovery Prediction Without a Vulnerability Database

Predicting software vulnerability discovery trends can help improve secure deployment of software applications and facilitate backup provisioning, disaster recovery, diversity planning, and maintenance scheduling. Vulnerability discovery models (VDMs) have been studied in the literature as a means to capture the underlying stochastic process. Based on the VDMs, a few vulnerability prediction schemes have been proposed. Unfortunately, all these schemes suffer from the same weaknesses: they require a large amount of historical vulnerability data from a database (hence they are not applicable to a newly released software application), their precision depends on the amount of training data, and they have significant amount of error in their estimates. In this work, we propose vulnerability scrying, a new paradigm for vulnerability discovery prediction based on code properties. Using compiler-based static analysis of a codebase, we extract code properties such as code complexity (cyclomatic complexity), and more importantly code quality (compliance with secure coding rules), from the source code of a software application. Then we propose a stochastic model which uses code properties as its parameters to predict vulnerability discovery. We have studied the impact of code properties on the vulnerability discovery trends by performing static analysis on the source code of four real-world software applications. We have used our scheme to predict vulnerability discovery in three other software applications. The results show that even though we use no historical data in our prediction, vulnerability scrying can predict vulnerability discovery with better precision and less divergence over time.

An expert system for channel routing

constrained approach to arrive at an acceptable routing. Some limitations of these approaches are discussed next. In this paper we present a Knowledge-Based system for multi-layer channel routing. Our system uses the blackboard model of architecture and is developed so as to exploit the shared-memory multiprocessor machines. This system utilizes a constraint posting approach. Unlike existing systems, it views the number of available layers as yet another design constraint on the channel being routed, hence rendering it free from layer dependence, a feature that all existing systems suffer from. A number of important routing metrics such as: 100% routabilify, minimum routing area, minimum wire length and minimum number of vias are considered simultaneously. The system has been implemented in Prelog and runs on Sequent Balance 8000. We ran several examples including the difficult ones and the results obtained are better than those obtained by existing routers that do not consider all the routing metrics. Out of the several routing-metrics, current algorithms concentrate on only one or at the most two of these metrics. The reason being that it is difficult to formulate a few heuristics that can guide routing and at the same time consider all the reutingmetrics. Absence of domain specific knowledge not only results in extensive search and backtracking but also deterioration of quality. Some algorithmic approaches allocate separate layers for the horizontal and vertical segments of nets. This constraint almost always leads to a larger routing area, longer interconnection wires and more vias. Most algorithmic approaches do not allow any obstruction in the routing area. Hence, interactive or manual routing of some critical nets is ruled out.

Analysis of the security of VPN configurations in industrial control environments

Abstract Virtual private networks (VPNs) are a popular approach for protecting otherwise insecure industrial control protocols. VPNs provide confidentiality, integrity and availability, and are often considered to be secure. However, implementation vulnerabilities and protocol flaws expose VPN weaknesses in many industrial deployments. This paper employs a probabilistic model to evaluate and quantify the security of VPN configurations. Simulations of the VPN model are conducted to investigate the trade-offs and parameter dependence in various VPN configurations. The experimental results provide recommendations for securing VPN deployments in industrial control environments.

Real-Time Vector Quantization and Clustering Based on Ordinary Differential Equations

This brief presents a dynamical system approach to vector quantization or clustering based on ordinary differential equations with the potential for real-time implementation. Two examples of different pattern clusters demonstrate that the model can successfully quantize different types of input patterns. Furthermore, we analyze and study the stability of our dynamical system. By discovering the equilibrium points for certain input patterns and analyzing their stability, we have shown the quantizing behavior of the system with respect to its vigilance parameter. The proposed system is applied to two real-world problems, providing comparable results to the best reported findings. This validates the effectiveness of our proposed approach.

Fast Fusion of Medical Images Based on Bayesian Risk Minimization and Pixon Map

Fast fusion of multiple registered out-of-focusimages is of great interest in medical imaging; forexample, the thoracic cavity is always too bumpy to befocused on all parts at one shot even when we can omitthe unavoidable hardware vibrations. Previousproposed methods in this field cannot fulfill the realtimerequirement in our multiple camera medicalimaging setting. In this paper, we propose a multiresolutionBayesian risk minimization based method tofuse these chest cavity images. The validity andefficiency of our method are verified by ourexperiments on both out-of-focus medical images andregional motion blurred images. By choosing specialkernel functions for the Pixon map and adoptinguniform distribution as the prior probability, ourmethod can be applied to the real-time medicalimaging situations such as surgical operationmonitoring.